Paul Vincent Craven
All Rights Reserved
Any sophisticated power management system will require information about
how the car is running and what kind of land it is driving on. The first
step in this process is to gather information from electronic sensors on
the solar car. In order to do this, the UMR team designed an embedded computer
to gather information about the car's operations. The computer in the Sunrayce
'95 car took in sensor information using two top-of-the-line A/D (analog-to-digital)
converter chips. These converters took the sensor information and converted
it to a digital format suitable for computer processing.
Figure 4: John Tyler and Dan Bohachic working on electronics
Figure 5: The Lead Van
Aside from the UPS loaned by Rock Mechanics, we also had one loaned from Midwest Computer Sales (a local store in Rolla). That UPS did not always like the generator power, however. We were guessing that the generator did not run a perfect 60 Hz sine wave, and the UPS did not like that. The voltage was correct, but we had no way to measure the frequency during the race. The UPS's primary purpose is to filter the generator power to a clean sine wave. Additionally, it keeps the equipment running if the generator stops. The UPS would keep the systems running until we could stop and refuel.
For Sunrayce '97, it is suggested to have either a larger engine alternator with an inverter to provide 110 volts, or mount the generator on the roof. Having the generator on the back of the vehicle blocked access to the rear doors, and produced a risk of carbon monoxide poisoning if the back door or window were left open. In addition, had the solar car rear-ended the lead vehicle it would have first hit the running generator and a gasoline container. Near the end of the race, several of us were watching the van as it sat with the generator running. Suddenly, one of the bolts holding it on failed, and half of the rack fell to the ground. Had we been moving with the solar car behind us, the results would not have been as insignificant as they were when the van wasn't moving. A larger alternator with an inverter providing 110 volts would eliminate the need to maintain a generator and provide a reduction of noise pollution. This solution would be ideal, if the alternator could provide enough current.
One of the computers in the lead van was a desktop model which could run either the Linux or MS-DOS operating system. It was not used much in day-to-day operations. The other two computers were laptops running Windows NT. Running under the operating system is a service. A service is a program that runs invisibly in the background of the operating system. For those familiar with UNIX, it is the Windows NT version of a daemon. This service receives the car data off the communications port from the Fluke modem receiving it. The service takes the data, processes it, and makes it available to any computer on the local network via a pipe. A pipe is a channel that can pass data between computers. With this set up we can have more than one person looking over the car data at the same time. It also provides a backup if one computer goes down.
Accessing the named pipe with the car data on it is a multi-threaded Windows NT program. This program can graph and display all of the sensor values in the car. The program can alert the personnel to any sensor value out of range. For example: drawing too much current, low battery voltage, excessive motor or brake temperature, etc. Integrated into this program is also the GPS data, which is described later in this thesis.
We found Windows NT to be very reliable, although our notebooks often couldn't take the day's heat and would "blue-screen" Windows NT with hardware errors. When NT gets a hardware error, it lets the user know with a blue screen that describes the error. The operating system is not able to recover from the types of errors that are displayed on the blue-screens, forcing a reboot. The notebooks worked well as long as we always ran the air-conditioning and pointed clip-on fans at the notebooks. The Liquid Crystal Display (LCD) screens often washed-out in the higher heat, and the monitors we brought were unusable in the glare of direct sunlight. Next time we will look for donations of desktop systems (which should be better in the heat) and anti-glare monitors. With the desktop's components spread out over a wider area than a notebooks, they should survive better in the heat.